First thing is to realize you are not building a sword, you are drawing a representation of a sword. So, if it can’t be seen, don’t model it.
Whether to separate things into different objects (hinge, screw heads, pivot pin) depends on how easy it is to model it separately or as part of the same object. All you see of the pivot pin is the head, which is the same diameter as the cylinder which holds it, so it can easily be modeled as an extension of that cylinder. You might find the screw head is easier to model as a separate object, placed in an indent in the hinge plate (and in that case, you make the screw heads child objects of the hinge parent.)
Another thing to watch out for are Blender default numbers: they are way too high. Rather than adding a circle to begin modeling a screw head with it’s default 32 vertices, try it with twelve, or, if you are using subsurf, 6 or 8, and see whether you can detect any difference in the rendered object. Frequently, you’ll find you can use many fewer vertices than the default number.
Basically, you add as much detail as you need to make your project look good, and no more.
This is the question i was looking for an answer too as well. I looked at a video of slow motion Call of Duty:AW weapons and i can see how they get their nice FPS with decent looking weapons, a lot of the details come through the normal map. I’m a total blender noob so i’ve only just understood what the normal map really is and it’s pretty awesome
Am i right in assuming this; If an object cant be modelled without breaking up the existing topology/makes the model much more complex just for that one piece then its worth having it has a separate model and making it a child of the main model?
That’s what I’ve been doing, too for the locomotive I’m working on. Just make sure you don’t accidentally hit “Remove Doubles” at the wrong time.
I usually build each section of something (like a door with hinges and screws in the hinges) as a separate object and then join them all together. It can really screw things up, though, if I hit “Remove Doubles” after joining a hinge and screw heads. But, if I ever do this, I just select the edge-loop they have in common and hit ‘v’ to separate them again.
From certain angles, you’d be able to see inside the blade through those transparent textures. If you can assure that the camera will never point in those directions, then go for it. Otherwise, to model a jagged edge, you actually need jagged geometry.
Ah okay, that definitely helps me a lot as i was stuck trying to model a Thompson M1a1 gun as i thought all my details on the side had to come out from the original base piece of geometry.
Apologies for this i hope im not hijacking here but another question - If say i had a a screw that went through a piece of geometry so it could be seen at either side of it would that cause any issues if that model was to be imported into a game like that?
True. Point taken, but most of the time, the camera won’t be close enough to the blade for this to be a problem. Also, the blade is most likely to be in fairly fast motion when visible which would make it almost impossible to see.
WOW thank you so much guys, very informative. :evilgrin:
so model what i need.
it doesn’t need to be joined
i have also noticed something strange happening to my mesh/renders if you look at the section right where the cursor is the material isn’t rendering properly, is this due to a problem with my topology or the heavy creases i am using, i have noticed this diminishes when i increase the sub divisions.
The more Geo the longer the render. IF this is a still 1 frame, not that big of deal to have many GEO faces, verts, etc. IF this is ever going for game or motion… Learn baking textures. NIGHT and DAY. Yes you can get a bump out of a texture, but not like a nick in the sword. More subtle like indentations or especially bolt heads and screws as a texture.